Information about Sub-Saharan Africa África subsahariana
Journal Issue
Share
Chapter

5. Global Value Chains

Author(s):
Céline Allard, Jorge Canales Kriljenko, Jesus Gonzalez-Garcia, Emmanouil Kitsios, Juan Trevino, and Wenjie Chen
Published Date:
March 2016
Share
  • ShareShare
Information about Sub-Saharan Africa África subsahariana
Show Summary Details

Beyond the pure expansion of trade, an additional dimension of globalization over the past two decades has been the emergence of global value chains (GVCs). In an increasingly integrated world economy fueled by technological progress, cheaper transportation and communication costs, and policy reforms in support of trade, production processes have been more dispersed across the globe. This has given rise to systems of supply chains in which value is added at each stage before crossing the border to be passed on to the next stage—GVCs. This process has allowed countries to better exploit their comparative advantages, by giving them the opportunity to join a production chain without having to provide all the other upstream capabilities, and has been particularly at play in South and East Asia around Japan and China and in Eastern Europe around Germany (IMF 2013, 2014b, 2015b; Chapter 3 of IMF 2014a).

For countries with a limited existing manufacturing or service export basis and a large pool of labor such as in those in sub-Saharan Africa, this development can provide a golden opportunity. By specializing on a specific segment of a production chain, each participating country can generate a portion of the goods’ or services’ value added—while producing the whole product from scratch would never have been within reach in an increasingly competitive world—even if that means that a lower share of the value added of exports is captured locally. While certain preconditions such as sufficient levels of capacity, quality, and efficiency are required to join GVCs (Baldwin 2011; WTO 2014), these threshold levels can be exceeded over time through technology and knowledge transfers from other countries—most often in the form of foreign direct investment (FDI). Furthermore, knowledge transfers from other producers in the value chain, and, eventually, upgrading to higher value-added segments of the production chain can support productivity and income growth. Asian countries have championed this model, initially contributing to the most labor-intensive activities in the production process and gradually moving into more sophisticated portions of the value chain.

To measure a country’s extent of international integration in GVCs, it is necessary to know the sources and destinations of the value added embodied in the products. A budding literature on trade in value added has emerged that relies on data using intercountry input-output (IO) tables. Until very recently, the coverage on sub-Saharan African countries in IO tables was sparse. We use here the newly created Eora database, which provides global multiregion IO tables, to derive value-added trade for 189 countries from 1990 to 2012 (Lenzen and others 2012; Lenzen and others 2013). The main advantage of using the Eora database is the depth of its coverage, in terms of countries (189), industries (about 16,000), and years (23 years); it is virtually unmatched by any existing database. It covers 42 out of the 45 countries in sub-Saharan Africa. While this extended coverage makes the database invaluable for the analysis conducted here, it should be remembered, however, that some missing data in the IO tables are filled through optimization procedures using existing national and global statistics: this means that our results should not be taken as exact and precise measures, although we believe the gist of the results to be robust.

The literature traditionally decomposes exports into three distinct components, which are used to measure GVC participation8:

  • Foreign value added (FVA) that has been imported from foreign suppliers upstream in the GVC. This share is referred to as backward integration, and reflects the extent to which a country is integrated relatively downstream of the value chain.
  • Domestic value added (DVA) of products consumed directly in the country where it is exported.
  • DVA of products that enter themselves into the production of other countries’ exports. This share is referred to as forward integration, and reflects the extent to which a country is integrated relatively upstream of the value chain.

The sum of the last two components corresponds to the DVA, and contributes toward its GDP. The sum of FVA and DVA results in the total value of gross exports. Figure 13 provides an illustrative example of a hypothetical value chain production of a shoe dispersed in different countries.

Figure 13.Value-Added Trade and Global Value Chains – An illustrative Example

Source: IMF staff.

Note: DVA is domestic value added in exports; FVA is foreign value added in exports.

The integration into GVCs has indeed gone hand in hand with a pickup in income levels. In particular, we focus on the measure of backward integration; that is, the FVA that is imported for further processing into exports. By this measure, rising backward integration has been associated with rising income over time for developing and emerging economies (Figure 14a). In pursuing a strategy of development anchored around integration in one intermediary link of the value chains, many countries have managed to lift their income levels as they gradually acquired new capabilities, and have benefited from knowledge spillovers and, eventually, from opportunities to diversify production and upgrade quality (UNCTAD 2013). In addition, enhanced participation in GVCs has also been associated with more inclusive growth, especially when the sectors targeted are labor-intensive and employ relatively lower-skilled workers.9

Figure 14a.Depth of Integration in Global Value Chains and Real GDP per Capita, Average 1991–95 and 2008–12 (full country sample)

Sources: World Bank, World Development Indicators; Eora database; and IMF staff calculations.

Where do sub-Saharan African countries stand in that landscape? Using the Eora multi-regional input-output database mentioned earlier, we can provide here a first-time assessment of the region’s positioning in GVC.

Sub-Saharan African countries still generally find themselves at the start of their integration process into GVCs, having also relatively lower income levels than other regions in the world (Figure 14b). At 15 percent of exports, the share of foreign value added embedded in the production of exports is low even compared with the 20 percent average observed in developing and emerging market economies. More worrisome is that the depth of its integration has barely increased since the mid-1990s, unlike in other income groups—signaling that the region has yet to join this global momentum and take advantage of it to lift productivity and create jobs (Figure 15). Corroborating that finding, neither the complexity of sub-Saharan African exports—measured as the diversity of products (Hausmann and others 2011)—nor the quality of exported goods—derived from price differences within specific product categories (Henn, Papageorgiou, and Spatafora 2013)—have been improving over the past two decades. In addition, compared with all other regions in the world, sub-Saharan African exports tend to enter at the very beginning of GVCs (in the form of forward integration), as a higher share of its exports enter as inputs for other countries’ exports, reflecting the still-predominant role of commodities in many countries’ exports in the region.

Figure 14b.Depth of Integration in Global Value Chains and Real GDP per Capita, Average 1991–95 and 2008–12 (subset of countries with 2005 GDP per capita below 3,500 U.S. dollars)

Sources: World Bank, World Development Indicators; Eora database; and IMF staff calculations.

Figure 15.Global Value Chains Participation, Average 1991–95 and 2008–12

Sources: Eora database; and IMF staff calculations.

1 Excluding sub-Saharan African countries.

There is, however, a significant degree of heterogeneity across sub-Saharan African countries, with some countries having fared much better than others (Figure 16):

  • Oil exporters are the least integrated in GVCs in terms of FVA content of their exports. With the exceptions of Cameroon and Congo, this share has even decreased, including in countries such as Angola and Nigeria, suggesting that diversification of trade away from natural resources has stagnated, if not gone backward, over the past 20 years in these countries.
  • However, in the rest of the region, a majority of countries (24 out of 35) have made progress, even if from a low starting point (Figure 17). The improvement is most widespread among non-oil-commodity exporters, with countries such as Burkina Faso, Central African Republic, Democratic Republic of the Congo, Ghana, Guinea, Niger, Sierra Leone, and Zimbabwe all registering progress. This shows that integration in value chains can happen even in countries where commodities play a role.
  • Among the best performers, progress within the EAC has been particularly strong, with Kenya, Tanzania, and Uganda exhibiting solid progress—also a reflection of the benefits of the more general economic integration at play among these countries and their stated intention to further deepen their economic and monetary ties (IMF 2015c; Sutton 2012). Likewise, the SACU region exhibits relatively stronger depth of integration, both because its smaller members (Botswana, Lesotho, Namibia, Swaziland) were already quite integrated in the early 1990s and because South Africa did progress over the 1990–2010 period. Conversely, both the CEMAC and the WAEMU continue to exhibit low depth of integration. For the former, this has to do with the high reliance on oil exports for most of its members.

Figure 16.Sub-Saharan Africa and Comparator Countries: Depth of Integration in Global Value Chains, Average 2008–12

Sources: Eora database; and IMF staff calculations.

Figure 17.Depth of Integration in Global Value Chains, Average 1991–95 and 2008–12

Sources: Eora database; and IMF staff calculations.

Note: See Annex 3.2 Country Groups for a list of countries in each group.

For the latter, this suggests that the relatively high level of interregional trade with the currency union does not reflect the emergence of a regional value chain, but rather trade on final goods and services, with the depth of integration particularly low for the two largest countries of the union—Côte d’Ivoire and Senegal.

  • Five countries in particular stand out, having seen the share of FVA in their exports increase by 5 percentage points or more in the past two decades: Ethiopia, Kenya, Seychelles, South Africa, and Tanzania (Figure 18). In these countries, the sectors that have benefited the most from the deepening of integration include agriculture and agro-business (especially in Ethiopia and Seychelles), and manufacturing (particularly in Tanzania), but also textiles, transport, and tourism, although to a lesser extent. These experiences bode well for the region: for one, the increase in depth of integration in some of these countries, at 10 percentage points or more, is of a similar magnitude to that experienced by countries such as Poland or Vietnam that are now success stories within large GVCs. The examples also highlight the sectors—agro-business, light manufacturing, tourism, and textile—in which sub-Saharan Africa has the potential to leverage its comparative advantages.
  • However, to leverage these comparative advantages, the business environment (infrastructure, rule of law, cost and wage competitiveness, and so on) needs to be right. On that front, more still needs to be done, judging from the broader trend decline in industrialization in the region documented in other studies (Rodrik 2015; Figure 19). It should be noted, though, that opportunities to participate in GVCs are not limited to manufacturing. Just as the production of goods has been broken down into different stages, services are increasingly being disaggregated and traded as separate tasks to create service value chains—as championed by India, for example.

Figure 18.Sub-Saharan African Selected Countries: Decomposition of Change in Depth of Integration in Global Value Chains, Average 1991–95 to 2008–12

Sources: EORA database; and IMF staff calculations.

1 Includes electrical and machinery, metal products, wood and paper, transport equipment, and other manufacturing.

2 Includes construction, telecommunications, wholesale trade, maintenance and repair.

Figure 19.Share of Manufacturing Value Added

Sources: Groningen Growth and Development Center database (Timmer, de Vries, and de Vries 2014); and IMF staff calculations.

Note: The figure depicts the share of manufacturing value added as a percent of the economy’s tota value added in 1995 on the x axis and that in 2010 on the y axis. The majority of the countries experienced decreases in teh share of manufacutring value added between the two years, depicted by the location underneath the 45-degree line, thus, inducating output deindustrialization (Rodrik 2015).

The upshot is that the region still has an enormous potential to integrate into GVCs. By leveraging this potential, a better insertion in GVCs may help foster structural transformation, export diversification, and the possibility to absorb technology and skills from abroad. These benefits are especially important for countries with relatively small domestic markets, such as many in sub-Saharan Africa; in addition, the enabling of strong job creation would also allow countries to harness the dividends of the upcoming demographic transition (see IMF 2015c).10

An additional question would be which country or region could serve as an anchor for sub-Saharan Africa’s integration into GVCs. Some larger and more advanced economies within the region, most notably South Africa, could be candidates. Alternatively, given growing ties with China and India, including through FDI, these emerging markets could see increasing value in outsourcing some of their economic activities to sub-Saharan Africa, especially as rising wages in the Asian countries could make the region more cost-competitive.

In that context, an econometric analysis investigates the policy measures likely to support a stronger insertion for the region into GVCs. We mimic the gravity equation in the previous section in terms of the control variables, but instead of bilateral trade, we use backward integration as the dependent variable:

where FVAit is the share of FVA in country i’s exports in year t, and where Xt–1 are control variables. For the latter, we use GDP per capita as well as the same term in squared term to capture the negative portion of the relationship. For policy variables, we include domestic credit provided by the financial sector as share of GDP, spending on education as share of GDP, quality of infrastructure, the weighted average of tariff rates applied to all products in a given country and year, and, last, a measure on the rule of law. All variables, with the exception of index variables, are in logs to eliminate potential outliers, and they are lagged by one year to avoid simultaneity bias. Moreover, we control for time, θt, and country, γi, fixed effects. As a robustness check, we also run a separate regression using the subsample of countries and years with only $22,000 GDP per capita, thus capturing only the portion in which backward integration and income levels are positively related, as evidenced in Table 3. The variables show similar magnitudes and levels of statistical significance.

Table 3.Panel Regression of Backward Integration and Policy Variables
Dependent Variable: Backward Integration as Share of Total Exports
(1)(2)
EntireCapita <
Sample$22,000
Real GDP per capita (lag1)0.326**−0.085*
(0.161)(0.047)
Real GDP per capita^2 (lag1)−0.029**
(0.011)
GDP (lag1)−0.060***−0.059***
(0.017)(0.020)
Domestic credit to private sector
(percent of GDP) (lag1)0.082*0.080
(0.043)(0.054)
Education (percent of GDP) (lag1)0.413***0.349***
(0.081)(0.082)
Rule of law (lag1)0.287***0.328***
(0.063)(0.063)
Quality of infrastructure (lag1)0.0470.063
(0.041)(0.048)
Tariff_weighted (lag1)−0.296***−0.254***
(0.037)(0.041)
Constant−2.672***−1.216***
(0.636)(0.380)
Number of observations385236
Time fixed effectsYesYes
Country fixed effectsYesYes
R-Squared0.390.57
Source: Eora database; World Bank, World Development Indicators; Global Competitiveness Index; and IMF staff calculations.Note: All variables are in natural log, except for rule of law and quality of infrastructure; the independent variables are lagged by one year. Robust standard errors in parentheses; * indicate significance at 10 percent, ** at 5 percent, and *** at 1 percent.
Source: Eora database; World Bank, World Development Indicators; Global Competitiveness Index; and IMF staff calculations.Note: All variables are in natural log, except for rule of law and quality of infrastructure; the independent variables are lagged by one year. Robust standard errors in parentheses; * indicate significance at 10 percent, ** at 5 percent, and *** at 1 percent.

We conduct the estimation on an unbalanced panel for 185 countries and over the period 2007–11. After controlling for the level of development and the size of the economy (as smaller countries tend to be more internationally integrated, all else equal), deeper integration in GVCs—as measured by a higher share of FVA in one country’s exports—is found to be associated with improved indicators of human capital and availability, while it is hampered by higher tariff levels and difficult business environments (Table 3). More specifically, a reduction in tariff rates across sub-Saharan Africa toward the average prevailing in non-sub-Saharan African countries could increase the share of FVA in exports by about 3 percentage points, an increase in access to credit by 2 percentage points, and an increase in education spending and rule of law to levels seen elsewhere in the world by another 1 percentage point each. While such changes would likely occur over time, together they would bring the depth of integration of the region to levels currently seen in other low-income and emerging markets. This suggests that actions on these policy levers would go a long way to positioning the region well to participate in GVCs.

    Other Resources Citing This Publication